---
title: "Estimating the burden of COVID-19 in African countries"
output: 
  flexdashboard::flex_dashboard:
    vertical_layout: fill
    orientation: rows
    source_code: embed
    theme: simplex
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(message = FALSE, warning = FALSE)

# Load Packages
library(highcharter)
library(htmltools)
library(rio)
library(stringr)
library(DT)
library(dplyr)
library(jsonlite)
library(data.table)
library(leaflet)

```


```{r ests, include = FALSE}
# read in data
country <- fromJSON("output/geojson/country.geojson", simplifyVector = FALSE)
admin <-  fromJSON("output/geojson/admin2.geojson", simplifyVector = FALSE)
country_dt <- fread("output/country_dt.csv")
admin_dt <- fread("output/admin2_dt.csv")

# get cfrs
# Per Jess
age.upper <- c(9, 19, 29, 39, 49, 59, 69, 79, 89)
N.cases <- c(416, 549, 3619, 7600, 8571, 10008, 8583, 3918, 1408)
N.deaths <- c(0, 1, 7, 18, 38, 130, 309, 312, 208)
CFR <- N.deaths/N.cases
fit.cfr <- smooth.spline(age.upper - 4.5, (N.deaths/ N.cases)) # mid-point of age bracket & cases/deaths

# take midpoint of our age brackets for afr data
age_lower <- seq(0, 65, by = 5)
age_upper <- age_lower + 4 # keep age upper at 69 (conservative estimates of mortality)
mid_pt <- (age_lower + age_upper)/2
cfr <- pmax(predict(fit.cfr, mid_pt)$y, 0)
names(cfr) <- c("A0004", "A0509", "A1014", "A1519", "A2024", "A2529", "A3034", "A3539", "A4044",
                "A4549", "A5054", "A5559", "A6064", "A65PL") # names should match colnames!
p_infected <- 0.3 # 30% cummulative infections

# Relative symptomatic by age (taking mid_pt of age bins)
cm_interpolate_cos <- function(x, x0, y0, x1, y1) {
  ifelse(x < x0, y0, 
         ifelse(x > x1, y1, y0 + (y1 - y0) * (0.5 - 0.5 * cos(pi * (x - x0) / (x1 - x0)))))
}
symp.pars <- data.frame("age_y" = 14, "age_m" = 55, "age_o" = 64, "symp_y" = 0.056,
                      "symp_m"= 0.49, "symp_o"= 0.74)
young  = cm_interpolate_cos(mid_pt, symp.pars$age_y, 1, symp.pars$age_m, 0);
old    = cm_interpolate_cos(mid_pt, symp.pars$age_m, 0, symp.pars$age_o, 1);
middle = 1 - young - old;
rel.symp <- young * symp.pars$symp_y + middle * symp.pars$symp_m + old * symp.pars$symp_o
names(rel.symp) <- c("A0004", "A0509", "A1014", "A1519", "A2024", "A2529", "A3034", "A3539", 
                     "A4044", "A4549", "A5054", "A5559", "A6064", "A65PL") # names should match colnames!

# apply to data @ country/admin level
admin_dt[, (paste0("deaths_", names(cfr))) :=  Map("*", .SD, cfr*p_infected*rel.symp), 
                           .SDcols = names(cfr)]
admin_dt$deaths_total <- rowSums(admin_dt[, paste0("deaths_", names(cfr)), with = FALSE], 
                                   na.rm = TRUE)
admin_dt$prop_ov65 <- admin_dt$A65PL/admin_dt$pop
admin_dt$inc_per100k <- admin_dt$deaths_total/admin_dt$pop*1e5

# country level
country_dt[, (paste0("deaths_", names(cfr))) :=  Map("*", .SD, cfr*p_infected*rel.symp), 
                           .SDcols = names(cfr)]
country_dt$deaths_total <- rowSums(country_dt[, paste0("deaths_", names(cfr)), with = FALSE], 
                                   na.rm = TRUE)
country_dt$prop_ov65 <- country_dt$A65PL/country_dt$pop
country_dt$inc_per100k <- country_dt$deaths_total/country_dt$pop*1e5

```


Sidebar {.sidebar}
======================================================================
Here we'll put info & links to other things...

Mapping burden
======================================================================

Row {{data-height=1000}}
-----------------------------------------------------------------------

### Burden mapped to country & Admin 2 level (where available)

``` {r, include = FALSE}

# create index vector to subset countries by admin level
country_dt$iso_id <- as.numeric(factor(country_dt$iso))
country_dt$value <- country_dt$inc_per100k
admin_dt$iso_id <- country_dt$iso_id[match(admin_dt$iso, country_dt$iso)]
admin_dt$value <- admin_dt$inc_per100k

cn_st_index <- unlist(lapply(admin$features, function (x){x$properties$iso}))
cn_st_index <- country_dt$iso_id[match(cn_st_index, country_dt$iso)]

# keep only what you need
admin_dt <- select(admin_dt, iso, country, admin_name, admin_type, 
                   pseektrthc10x10, pop, prop_ov65, value, deaths_total, id_match = shape_id, iso_id)

build_series <- function(subnatl) {
  
  # subset uscountygeojson
  admin.subset <- admin
  admin.subset$features <- admin$features[which(cn_st_index == subnatl)]
  
  # subset county data
  ds.cn <- filter(admin_dt, iso_id == subnatl) %>%
    list_parse()
  
  # build series
  list(
    id = subnatl,
    mapData = admin.subset,
    data = ds.cn,
    joinBy = c('id_match', 'id_match'),
    dataLabels = list(enabled = TRUE, format = '{point.name}'),
    tooltip = list(
      useHTML = TRUE,
      headerFormat = "
",
      pointFormat = paste0("{point.country}
",
                           "{point.admin_name} ({point.admin_type})
",
                           " Pop : {point.pop:.0f}
",
                           " Prop pop over 65: {point.prop_ov65:.3f}
",
                           " Deaths (total #): {point.deaths_total:.0f}
",
                           " Estimated reporting of fevers to health center: {point.pseektrthc10x10:.2f}
"),
      footerFormat = "
")
  )
}

# Make the graph
# keep only neccesary data!
country_dt %>% 
  select(iso, country, pop, prop_ov65, iso_id, value, deaths_total, pseektrthc10x10) %>%
  rename(drilldown = iso_id) %>%
  mutate(pop_m = pop/1e6) -> dt.st
ds.st <- list_parse(dt.st)

# create drilldown series
series.list <- lapply(as.numeric(dt.st$drilldown), build_series)

```

```{r}
  # tooltip for country map has to be coded separately from tooltip for county
  # max inc
  max_inc <- max(max(country_dt$value, na.rm = TRUE), max(admin_dt$value, na.rm = TRUE))
                 
  highchart(type = 'map') %>%
    hc_add_series(
      mapData = country, data = ds.st, joinBy = c("iso", "iso"),
      borderWidth = 0.8, dataLabels = list(enabled = FALSE, format = '{point.properties.iso}'),
      tooltip = list( 
        useHTML = TRUE,
        headerFormat = "
",
        pointFormat = paste0("{point.country}
",
                             " Pop : {point.pop_m:.3f} million
",
                           " Prop pop over 65: {point.prop_ov65:.3f}
",
                           " Deaths (total #): {point.deaths_total:.0f}
",
                           " Estimated reporting of fevers to health center: {point.pseektrthc10x10:.3f}
"
                           ),
        footerFormat = "
"
      )) %>%
    hc_colorAxis(stops = color_stops(9, c('#fff5f0', '#fee0d2', '#fcbba1', '#fc9272', '#fb6a4a', 
                                           '#ef3b2c', '#cb181d', '#a50f15', '#67000d'))) %>%    
    hc_legend(enabled = TRUE, title = "Incidence of deaths per 100k persons") %>%
    hc_exporting(enabled = TRUE)%>%
    hc_drilldown(
      series = series.list,
      activeDataLabelStyle = list(
        color =  '#FFFFFF',
        textDecoration = 'none'
      )
    ) %>%
    hc_mapNavigation(enabled = TRUE) %>%
    browsable()
```

Methods
======================================================================

Coming soon!